Hi,
I was wondering if it is possible to run DESeq ( or for that reason any other tool for differential expression analysis) to identify significant differences between two organisms (for our research human and mouse).
I know that we have two different sets of genes, but what if we can overcome this problem by giving the genes a neutral ID based on a common variable and take only the genes which are present in both data sets.
We have several RNA-Seq data sets with multiple samples (=replica) from different brain layers in mouse and human. We are doing a DE analysis (DESeq2) to find out which genes are significantly changing between the different brain layers. But we are also interested in changes in the same layer but between the two organisms.
How does it looks like from the statistical point of view?
any ideas?
thanks,
Assa
You'll need to tell us your experimental design. In general, what you are describing is not usually possible to do directly.
I have edited the question and hope it is a bit clearer now.
I was thinking that since DESeq2 needs "only" a count table, why not just take the human samples from layer A and compare them with the samples of layer A from the mouse.
For common probe names I was thinking about taking the gene symbol as a unified ID and than only take the genes which are in both data sets (not in the RNA-Seq results, but from the complete data set, e.g. gtf file).